Association Is Broader Than Linear Association
In “Recognizing Linear and Nonlinear Form,” you learned to distinguish a roughly straight pattern from one that bends. That distinction matters because association is the broad idea: paired values of two quantitative variables show a pattern in how they vary together. Linear association is a particular kind of association, one whose overall pattern is reasonably described by a straight line.
A relationship does not need to follow a straight line to be an association. If the points follow a clear curve, the variables are associated, even though the association is nonlinear. A close U-shaped pattern is a useful example: the response tends to decrease over part of the explanatory variable’s range, then increase over another part. That systematic change is a pattern, not a lack of one.
Keep the features you learned in “Describing a Scatterplot With DUFS” distinct. Form tells you whether the pattern is roughly straight or curved; strength tells you how closely the points follow the overall pattern; and direction describes how the response tends to change as the explanatory variable increases. These descriptions work together, but one does not replace the others. A curved pattern can be strong, and its direction can change across the graph.
For a U shape, moving from left to right, the response first tends to fall and then tends to rise. It would be misleading to describe the entire relationship as simply positive or simply negative. The relationship still has a clear form and can be strong: the points may follow the curve closely even though they do not follow one straight trend.
Why a U Shape Is an Association
Consider a scatterplot of outdoor temperature and a building’s daily energy use. Suppose energy use is low near a comfortable temperature but higher on much colder or much hotter days. The points might trace a U shape: moving toward the middle of the temperature range, energy use tends to fall; moving past the middle, it tends to rise.
This pattern is not well described by saying “higher temperature goes with higher energy use” or “higher temperature goes with lower energy use” across the whole range. Both statements might describe one part of the graph, but neither describes the complete pattern. A useful description names the bend and, when helpful, explains how the response changes on either side of the low point.
If the points lie close to the U-shaped curve, the association is strong in the sense introduced in “Judging Strength of an Association”: the observations follow the overall pattern closely. Strong does not mean straight. A straight-line description would miss the bend and could give a poor account of how energy use changes at different temperatures.
Worked Example: Temperature and Building Energy Use
Suppose a fictional building manager records outdoor temperature and the building’s daily energy use on seven days. These invented observations are for practice, not a real study. Temperature is the explanatory variable, measured in degrees Celsius; energy use is the response, measured in kilowatt-hours (kWh).
| Outdoor temperature (°C) | Daily energy use (kWh) |
|---|---|
| 5 | 46 |
| 10 | 26 |
| 15 | 14 |
| 20 | 10 |
| 25 | 14 |
| 30 | 26 |
| 35 | 46 |
Read the pattern from left to right. As outdoor temperature increases from 5°C to 20°C, energy use decreases from 46 kWh to 10 kWh. Above 20°C, energy use increases again, reaching 46 kWh at 35°C. The response therefore does not consistently rise or fall across the full range of temperatures.
Describe the form and strength. The values make a clear U-shaped pattern: energy use is lowest near 20°C and increases toward both cooler and warmer temperatures. The observations follow that pattern closely, so this is a strong association with nonlinear form. It is not a strong linear association; one straight trend would fail to represent the decrease followed by the increase.
Conclusion in context. In these invented observations, outdoor temperature and daily energy use have a strong, nonlinear, U-shaped association. Energy use tends to be lower near 20°C and higher at temperatures farther below or above it. The pattern alone does not establish what caused the differences in energy use.
Notice that the U shape includes two local tendencies: a negative direction on the left side, where energy use falls as temperature rises, and a positive direction on the right side, where energy use rises as temperature rises. Naming both parts is more informative than forcing a single positive or negative label on the whole pattern.
Curved Patterns Can Rise Without Being Linear
A U shape is not the only kind of nonlinear association. Some curved patterns rise throughout the observed range but rise at a changing rate. In that case, it can be accurate to describe the overall direction as positive while also saying that the form is curved rather than roughly straight. Direction and form answer different questions.
For example, a distance-versus-time scatterplot for an object speeding up may rise more steeply as time passes. The response, distance traveled, increases as the explanatory variable, time, increases. But if the points bend upward rather than following a straight path, the relationship is positive and nonlinear—not a linear association. The context must support the interpretation; a graph alone shows a pattern, not the mechanism that produced it.
Worked Example: Time and Distance for an Accelerating Cart
Imagine a fictional classroom demonstration in which students record the time and distance traveled by a cart that speeds up along a track. The values below are invented. Time is measured in seconds, and distance from the starting point is measured in meters.
| Time (seconds) | Distance (meters) |
|---|---|
| 0 | 0 |
| 1 | 2 |
| 2 | 8 |
| 3 | 18 |
| 4 | 32 |
| 5 | 50 |
Identify the overall direction. As time increases from 0 to 5 seconds, the distance from the starting point increases at every listed observation. The overall direction is positive.
Check whether the form is linear. The increases in distance over successive one-second intervals are 2, 6, 10, 14, and 18 meters. Those increases are not approximately constant, so the listed pattern bends upward rather than following a straight-line trend. The direction is positive, but the association is nonlinear.
Conclusion in context. In this fictional demonstration, elapsed time and distance traveled have a positive, curved association: the cart is farther from its starting point at later times, and the distance increases at a changing rate. Calling the pattern linear would overlook its upward bend.
A direction label by itself would not fully describe these data. “Positive association” tells the reader that distance tends to increase with time, while “nonlinear” tells the reader that a straight-line form does not adequately describe the pattern. Both details are needed for a clear description.
Compare the Pattern, Not Just the Direction
The next example contrasts a roughly straight pattern with the curved patterns above. Both can have a positive direction, but only the first is reasonably described as linear. A table is not a substitute for viewing a scatterplot, yet paired values can help you notice whether the changes look roughly steady or systematically bend.
Worked Example: Practice Time and a Skill Score
A fictional coach records practice time and a skill score for six training sessions by one participant. Practice time is measured in hours, and the score is measured in points. These invented observations show a roughly straight positive pattern.
| Practice time (hours) | Skill score (points) |
|---|---|
| 1 | 12 |
| 2 | 15 |
| 3 | 17 |
| 4 | 21 |
| 5 | 23 |
| 6 | 27 |
Describe the direction. Higher practice times generally go with higher skill scores, so the association is positive.
Describe the form. The scores increase at a broadly steady rate, with some small variation. There is no pronounced U shape or other systematic bend in these values, so a roughly straight pattern is a reasonable description. Unlike the temperature and energy-use example, this is a positive linear association.
Conclusion in context. For these six fictional sessions, practice time and skill score show a positive, roughly linear association: sessions with more practice time tend to have higher skill scores. This description does not show that practice time alone caused the scores to increase.
The comparison makes the key distinction visible: a positive direction can appear in a linear pattern or in a curved pattern. To classify form, inspect the overall shape rather than deciding from the direction alone.
A Practical Check for Curvature
When you inspect a scatterplot, do not stop after noticing that values generally rise or fall. Look for a systematic bend. Ask whether the response changes in a similar way across the full range of the explanatory variable, or whether the pattern turns, flattens, or becomes steeper. A bend shared by many points is part of the form, not an unusual feature to dismiss.
Name the explanatory variable on the horizontal axis and the response variable on the vertical axis, including units when available.
Decide whether the pattern is roughly straight or has a systematic bend, such as a U shape.
Move from left to right. If the response changes direction, describe what it tends to do on each part instead of assigning one direction to the entire pattern.
Consider how closely the points follow the overall straight or curved form. A close curve can indicate a strong association even when the association is not linear.
Name the variables and describe the form, direction when appropriate, and strength. Do not turn an association into a cause-and-effect claim.
This check builds directly on DUFS: direction, unusual features, form, and strength. It helps prevent a common mistake—treating “positive or negative?” as if it were the only question. With a U-shaped pattern, form is especially important because the direction changes across the explanatory variable’s range.
Common Mistakes and AP Exam Tips
- Calling a U-shaped pattern “no association” because there is no single direction: A clear bend is a systematic relationship. Say “strong nonlinear association” or “strong U-shaped association,” as appropriate.
- Calling every positive association linear: Positive describes direction; linear describes form. A pattern can rise overall and still curve. State direction and form separately.
- Giving a single direction for the entire U shape: On the left side, the response tends to fall as the explanatory variable increases; on the right side, it tends to rise. Describe the change in direction.
- Using “strong” to mean “straight”: Strength describes how closely points follow a pattern. Form describes whether that pattern is straight or curved. A strong curve is possible.
- Fitting the description to a line when the points clearly bend: A straight-line description would miss important structure in the data. First inspect the scatterplot’s form, as emphasized in “Recognizing Linear and Nonlinear Form.”
- Claiming the explanatory variable caused the response: A scatterplot describes association. Unless the study design supports a causal conclusion, describe what tends to occur together rather than claiming that one variable produced the other.
For full credit, do not merely write “there is a relationship.” Name the variables and describe the pattern in context. For the energy-use example, an appropriate statement is: “Outdoor temperature and daily energy use show a strong U-shaped, nonlinear association; energy use tends to be lowest near 20°C and higher at cooler and warmer temperatures.” This identifies the form, explains the changing direction, and avoids claiming that the plot proves a cause.
Check Your Understanding
Use the distinction between association, direction, form, and strength to answer each question.
- A scatterplot has points close to a U-shaped curve. Is it accurate to call the association strong? Is it accurate to call it linear? Explain both answers.
- As a plant’s age increases, its height rises quickly at first and then levels off. What direction does the pattern have, and is its form necessarily linear?
- In the building example, what tends to happen to energy use as temperature increases on the left side of the U? What happens on the right side?
- A student says, “The relationship is not linear, so there is no association.” Explain the mistake.
- Write a one-sentence contextual description for a curved, positive pattern between time and distance that identifies both direction and form without claiming causation.